Control device and method of operating the control device

Abstract

A control device contains a plurality of controller chips connected in parallel on a data side, a common signal output, and a change-over unit. The controller chips are coupled on an output side to the common signal output through the change-over unit. A plurality of differentiating elements are disposed upstream of the change-over unit on the data side and in each case associated with one of the controller chips. An integrating element follows the change-over unit. An output of a controller chip is first differentiated, sent through the change-over unit, and than integrated.

Description

CROSS-REFERENCE TO RELATED APPLICATION

This is a continuing application, under 35 U.S.C. § 120, of copending international application No. PCT/EP2005/012898, filed Dec. 02, 2005, which designated the United States; this application also claims the priority, under 35 U.S.C. § 119, of German patent application No. 10 2004 058 328.5-55, filed Dec. 02, 2004; the prior applications are herewith incorporated by reference in their entirety.

BACKGROUND OF THE INVENTION

Field of the Invention

The invention relates to a control device containing a multiplicity of controller chips connected in parallel on the data side, and which are connected on the output side to a common signal output via a change-over unit.

In industrial control devices, a multiplicity of controller chips connected in parallel on the data side is usually provided, for example for redundancy or safety reasons or also for providing a multiplicity of alternative open-loop or closed-loop control processes, for providing a control signal. These controller chips are connected at the output side to a common signal output via a change-over unit, wherein in each case one of the controller chips can be selectively selected via the change-over unit to provide the output signal. Depending on the process sequence, it is usually predetermined which of the controller chips is to be currently active in a respective operating state. Even the simplest control devices usually have at least two such control channels, predetermined by alternative controller chips, which can be selectively selected, wherein a distinction can be made, for example, between the two operating states “automatic control” and “manual control”. In such cases, it is also possible to predetermine a value by selective intervention from the outside by selectively selecting the “manual control” channel (also called open-loop control) as an alternative to automatic control. In addition, a large number of alternative control channels, which can differ, for example, with regard to the internal process sequences or individually predetermined operating parameters, within a control device. Finally, a change-over between controller chips can also be provided which in each case control different process variables.

During the change-over between different operating modes, in such a control device the controller chip which happens to be currently active is separated on the signal side from the signal output by operating the change-over unit and an alternative controller chip is connected on the signal output on the signal side in an associated manner. For operating reasons, however, it is not always possible to ensure for such a change-over, that the controller chip to be switched off and the controller chip to be newly added have exactly the same output signal at the change-over time since slight signal differences may occur due to the possible heterogeneity of the processes running in the controller chips. Such signal differences may even be comparatively large since the controller chip to be newly added is not in a closed control loop at the time of change-over and, therefore, may have the tendency of “running away” to a stop position. Signal deviations occur especially if different controlled variables are allocated to the controller chips. If such a signal deviation between controller chips occurs during the change-over, the output signal produces a so-called signal jump during the change-over. Such signal jumps can be very undesirable with regard to operating requirements or possible stability criteria.

To provide for a so-called smooth change-over in order to avoid such a signal jump during the change-over between various controller chips, which guarantees distinctly increased control-related reliability of the control device overall, the concepts of relieving or of balancing can be provided, in principle. In the case of relieving, outputs of the individual controller chips are not switched but forwarded to the output channel via a minimum or maximum element. In such a circuit, the change-over is always smooth. However, the application of such a concept is comparatively greatly restricted since the minimum or maximum chip at the signal output completely defines the change-over rule. Implementing a change-over in accordance with any desired rule which is neither minimum nor maximum is not possible in this arrangement. In addition, the concept of relieving is too slow in most cases of control applications since the “relieving” controller must pass through a “distance” until the relief can take place. For this reason, the concept of relieving is not widely used especially in industrial applications of control devices.

As an alternative, the so-called “balancing” can be used which allows a change-over in accordance with any criterion, formed by the so-called “change-over logic”. In such a circuit, the output signal of a first controller chip can be switched as necessary as an additional input signal to an alternative controller chip now to be switched and taken into consideration in the latter during this process. However, such a circuit requires the mutual optional application of the output signals of all other control chips to all control chips. For this reason, such a concept exhibits a particularly high complexity, especially when a multiplicity of control chips are connected in parallel, which can result in a particularly high production or planning expenditure.

In addition, the concept of balancing is associated with the disadvantage that a “zero signal” is output for a short time during the change-over especially when implemented in analog technology. Such a change-over is thus not smooth in the narrow sense, at least in an analog application.

In the case of an alternative application in a digital embodiment, in contrast, each individual balancing circuit represents a signal loop which cannot be timeless in digital technology. However, such signal loops have disadvantages in digital technology since it is not only the correctness of interconnection but also the variation with time of the balancing signals which must be taken into consideration especially in the planning. This problem can be made more difficult if a number of processors are used which run asynchronously with respect to one another. Thus, this concept can only be used to a limited extent especially in complex control systems.

BRIEF SUMMARY OF THE INVENTION

It is accordingly an object of the invention to provide a control device and a method of operating the control device which overcome the above-mentioned disadvantages of the prior art devices and methods of this general type, by which a smooth change-over between individual controller chips is made possible with particularly simple measures and in a reliable manner.

With the foregoing and other objects in view there is provided, in accordance with the invention a control device. The control device contains a plurality of controller chips connected in parallel on a data side, an output side having a common signal output, and a change-over unit. The controller chips are coupled to the common signal output through the change-over unit. A plurality of differentiating elements are disposed upstream of the change-over unit on the data side and in each case are associated with one of the controller chips. An integrating element is disposed downstream of the change-over unit.

With respect to the control device, the object is achieved, according to the invention, in that the change-over unit is preceded on the data side by a multiplicity of differentiating elements in each case allocated to one of the controller chips and followed by an integrating element.

The control device is thus configured for the outputs of the individual controller chips to be initially differentiated over time, then changed over and after that integrated over time in order to provide the output signal. The invention is based on the concept that, to provide a smooth change-over which is kept particularly simple, a permanent transverse balancing between individual controller chips to avoid signal jumps is to be prevented. Instead, the output signals of the controller chips should be processed in such a manner that, instead of the actual output signals possibly exhibiting individual deviations or jumps, a systematic arrangement common to all controller chips connected in parallel should be used. For this purpose, the derivation over time, which should have a comparable behavior due to the common system used for all controller channels, is first formed from the output signals delivered by the control chips. The change-over can then take place on the basis of this derivation over time of the actual signals and the actual output signal is restored after the change-over by forming the integral over time. The effects of differentiation and of the subsequent integration substantially cancel and the output signal does not exhibit a jump during the change-over since it is not the signal itself but its derivation which is switched. In the case of such a change-over, deviations between the individual controller channels or the controllers can only lead to a discontinuity in the derivation over time of the output signal, if at all, but this is not associated with a discontinuity in the actual signal.

In the configuration of the control device, an individual differentiating element kept separate can be allocated to each controller chip. However, one or each of the differentiating elements is advantageously integrated into its in each case associated controller chip. In this way, the controller chip including its associated differentiating element can be utilized as separate component in the manner of a modular configuration and connected directly to the subsequent change-over unit if necessary.

It is especially in the case of a PI controller (proportional integral controller), which is widely used per se, as a controller chip, that a controller chip provided with a differentiating element can advantageously comprise, on the one hand, a proportional branch provided with a proportional element and, on the other hand, a differential branch provided with a differentiating element, which are connected on the output side to a common summing element. In such an embodiment, the sum of a proportional proportion and a differential proportion of the control signal is thus formed within the controller chip. During the integration of the signal following the change-over, a signal with an integral component (formed from the proportional component previously contained) and a proportional component (formed from the differential component previously determined) is thus produced as output signal so that the conventional characteristics of the output signal of a PI controller are obtained.

With respect to the method, the object is achieved by supplying to the change-over unit a differential signal which is characteristic of the derivation with time of an output signal of a controller chip as input signal, from which, on the output side of the change-over unit, the output signal of the control device is generated by integration with time.

During this process, the differential signal is advantageously generated by derivation over time of the output signal of a controller chip. In an alternative, advantageous embodiment, the differential signal is formed by summing a first signal contribution formed by the derivation over time of a controller signal and a second signal contribution proportional to the controller signal in order to replicate a PI controller.

The advantages achieved by the invention particularly consist in that, due to the differentiation over time of the controller signals followed by integration over time, a substantially unchanged output signal is generated in the end effect, in which differentiation and integration cancel per se. Due to the change-over connected between the differentiation with time and the integration with time, a jump in the signal is reliably prevented during the change-over since it is not the signal but its derivation which is switched. This provides for an actually smooth change-over between the controller chips with particularly simple measures and without additional complexity in the interconnection of the controller chips with one another. Due to the comparatively simple concept, there are no further restrictions with regard to the type of controller chips used or of the change-over criteria. In particular, it is not required to equip the controller chips with a balancing function. In addition, no signal loops are required in the interconnection, the controller configuration being basically rapid, and arbitrary linear, nonlinear and/or adaptive controllers can be used as controller chips.

Other features which are considered as characteristic for the invention are set forth in the appended claims.

Although the invention is illustrated and described herein as embodied in a control device and a method of operating the control device, it is nevertheless not intended to be limited to the details shown, since various modifications and structural changes may be made therein without departing from the spirit of the invention and within the scope and range of equivalents of the claims.

The construction and method of operation of the invention, however, together with additional objects and advantages thereof will be best understood from the following description of specific embodiments when read in connection with the accompanying drawings.

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING

FIG. 1 is a block circuit diagram of a control device according to the invention;

FIGS. 2A and 2B are is a block circuit diagrams each showing a controller chip; and

FIG. 3 is a block circuit diagram of an alternative embodiment of a control device.

DESCRIPTION OF A PREFERRED EMBODIMENT OF THE INVENTION

Identical parts are provided with the same reference symbols in all figures. Referring now to the figures of the drawing in detail and first, particularly, to FIG. 1 thereof, there is shown a control device 1 that includes, in the manner of a multi-channel embodiment in which a multiplicity of control channels 2 are connected in parallel on the data side, a multiplicity of controller chips 4 which are connected in parallel on the data side and are connected on the output side to a common signal output 8 via a change-over unit 6. The change-over unit 6 is driven by change-over logic 10. Due to its multi-channel configuration, the control device 1 is particularly suitable for industrial applications in which alternately one of the controller chips 4 is to be active in each case in dependence on requirement and/or operating mode. It could be provided, for example, that one of the controller chips 4 is to be configured for manual intervention from the outside in the manner of manual control (also called open-loop control), wherein the remaining controller chips 4 can be configured for automated process control, possibly in accordance with sequence patterns deviating from one another or the like.

The control device 1 is configured for particularly high operating reliability and operational stability. This takes into account, in particular, that the controller chips 4, due to intrinsic characteristics or due to different controlled variables and control functions or due to the fact that controllers which are not engaged have a tendency to “run away”, could generate at their outputs output signals which differ from one another to a greater or lesser extent as a result of which it could come to more or less distinct signal jumps at the signal output 8 with a direct change-over between individual controller channels 2 or controllers. To avoid such unwanted signal jumps which, for example, could lead to excitation of unwanted or inadmissible oscillations in the control loop, the control device 1 is configured for so-called smooth change-over in which such signal jumps could especially be eliminated.

To achieve this by particularly simple measures, each control channel 2 in the control device 1 is in each case provided with a differentiating element 12 which is allocated to the respective controller chip 4 and is connected on the data side between the latter and the change-over unit 6. In addition, the change-over unit 6 is followed by an integrating element 14 at the data side.

In this circuit, the output signal RA output by the respective controller chip 4 in each controller channel 2 is first differentiated over time in the respective associated differentiating element 12 and during this process is converted into a differential signal D when the control device 1 is operated. The differential signal D is then applied to the change-over unit 6 where a change-over can occur between individual controller channels 2, if necessary. During this process, the change-over unit 6 forwards on on the output side from the controller channel 2 in each case selected as active the differential signal D fed in by the latter. This forwarded differential signal D is then integrated over time in the subsequent integration element 14. The result of this operation is delivered to the signal output 8 as output signal A. In this operating mode of the control device 1, the differentiation and subsequent integration of the output signal RA of the respective controller chip 4 cancel so that the output signal A generated by the control device 1 substantially remains unchanged by these operations. However, the change-over between the controller channels 2 in the change-over unit 6 is carried out not on the actual output signal RA of the controller chips 4 but on its derivation over time represented by a differential signal D. Thus, the deviations in the output signals RA of the controller chips 4 do not have any effect with respect to one another and the output signal A does not exhibit a jump during the change-over. The change-over is thus smooth in the narrower sense.

In principle, the concept based on the change-over of the derivation over time of the output signal RA of the controller chips 4 with integration with time taking place after the change-over does not necessitate any restrictions with regard to the type of controllers used or the change-over criteria. In particular, it is not required that the controller chips 4 to be changed over must have balancing functions or the like. In particular, the widely used controllers of the type P (proportional), I (integral), PI, PD (proportional differential), PID or also any linear, non-linear or adaptive controllers can be changed over smoothly with particularly little device and design expenditure, without requiring signal loops. In addition, the controllers between which change-over is effected do not even need to be of the same type. Instead, they can even control different process or controlled variables in each case. For example, one of the controllers could be a current controller and another controller could be a voltage controller.

The controller chips 4 can be PI (proportional-integral) controllers as is shown diagrammatically and illustratively in FIG. 2A. In this configuration, the controller chip 4 contains a proportional channel 20 and an integral channel 22 which is connected in parallel with the former on the data side, into which an integrating element 24 is connected. On the output side, the proportional channel 20 and the integral channel 22 are connected to a common summing element 26 in which the sum of the two part-signals is formed. This is followed at the output side by the differentiating element 12 which, in turn, is connected at the output side to the change-over unit 6 as indicated by the change-over point 28. With a corresponding “active” setting in the change-over unit 6, the controller channel 2 thus formed is connected to the subsequent integrating element 14. In the case of a configuration according to FIG. 2A, the controller chip 4 disposed in conventional PI construction is thus connected to the differentiating element 12, essentially kept separately, for forming the respective controller channel 2.

A system which is equivalent with regard to its control characteristics can be provided by integrating the differentiating element 12 into the controller chip 4 allocated to it in each case. An exemplary embodiment of such an arrangement is shown by a PI controller in FIG. 2B. The controller chip 4 provided with the differentiating element 12 contains, on the one hand, a proportional branch 32 provided with a proportional element 30 and, on the other hand, a differential branch 34 provided with a differentiating element 12. The outputs of the proportional branch 32 and of the differential branch 34 are connected to a common summing element 36. In this embodiment, the differential signal D characteristic of the derivation with time of the output signal RA of the controller chip 4 is thus formed by summing a first signal contribution S1, formed by the derivation over time of a controller signal, and a second signal contribution S2 proportional to the controller signal. The differential signal D thus output is supplied to the change-over unit 6 characterized via the change-over point 28 in accordance with the embodiment described above and is forwarded to the subsequent integrating element 14 in dependence on requirement and operating mode.

FIG. 3 shows an exemplary embodiment of a control device 1′ formed from such controller chips 4′ in each case already provided with an integrated differentiating element. Retaining the concept that the change-over between controller channels or controllers is to take place with regard to the derivation of the actual controller signal, wherein a subsequent integration with time is provided, the controller chips 4′, supplying a differential signal D as output signal in any case, are directly connected to the change-over unit 6 in this embodiment. The latter, in turn, is followed by the integrating element 14 for forming the actual output signal A.

Claims

1. A control device, comprising:

a plurality of controller chips connected in parallel on a data side;

an output side having a common signal output;

a change-over unit, said controller chips coupled to said common signal output through said change-over unit:

a plurality of differentiating elements disposed upstream of said change-over unit on said data side and in each case associated with one of said controller chips; and

an integrating element disposed downstream of said change-over unit.

2. The control device according to claim 1, wherein at least one of said differentiating elements is integrated into said associated controller chip.

3. The control device according to claim 2, wherein at least one of said controller chips with said respective differentiating element integrated therein further includes a common summing element, a proportional branch having a proportional element and a first output, and a differential branch having said respective differentiating element and a second output, said first and second outputs connected to said common summing element.

4. A method for operating a control device containing a plurality of controller chips connected in parallel on a data side, and coupled to a common signal output on an output side through a change-over unit, which comprises the steps of:

supplying the change-over unit with a differential signal being a characteristic of a derivation over time of a controller output signal of a controller chip as an input signal to the change-over unit; and

integrating over time the differential signal output by the change-over unit resulting in an output signal of the control device.

5. The method according to claim 4, which further comprises generating the differential signal by derivation over time of the controller output signal of the controller chip.

6. The method according to claim 4, which further comprises forming the differential signal by summing a first signal contribution formed by the derivation over time of the controller output signal, and a second signal contribution proportional to the controller output signal.